Mtr4-like protein coordinates nuclear RNA processing for heterochromatin assembly and for telomere maintenance

Abstract

The regulation of protein-coding and noncoding RNAs is linked to nuclear processes, including chromatin modifications and gene silencing. However, the mechanisms that distinguish RNAs and mediate their functions are poorly understood. We describe a nuclear RNA-processing network in fission yeast with a core module comprising the Mtr4-like protein, Mtl1, and the zinc-finger protein, Red1. The Mtl1-Red1 core promotes degradation of mRNAs and noncoding RNAs and associates with different proteins to assemble heterochromatin via distinct mechanisms. Mtl1 also forms Red1-independent interactions with evolutionarily conserved proteins named Nrl1 and Ctr1, which associate with splicing factors. Whereas Nrl1 targets transcripts with cryptic introns to form heterochromatin at developmental genes and retrotransposons, Ctr1 functions in processing intron-containing telomerase RNA. Together with our discovery of widespread cryptic introns, including in noncoding RNAs, these findings reveal unique cellular strategies for recognizing regulatory RNAs and coordinating their functions in response to developmental and environmental cues.

Figure 1. Red1 and Mtl1 Form a Common Core that Associates with Other Proteins

(A) Expression of tagged proteins. Extracts from tagged and untagged strains were analyzed by Western blot. (B) Proteins co-purified with Red1, Mtl1, Pir1, and Rmn1. Proteins highly associated with each other are shaded in blue. (C) Co-IP of associated proteins from strains expressing tagged proteins. (D) Immunofluorescence analysis of MYC-tagged proteins. (E) Co-IP analysis of Pla1 and Mtl1 interaction. (F) Protein interaction network of MTREC and Mtl1. MTREC, comprising Mtl1 and Red1, is the core module that interacts with Pir1, Rmn1-Pab2, or Pla1 to form different functional modules (left). Mtl1 also forms Red1-independent interactions with Nrl1 and Ctr1 (right). See also Figure S1.

Figure 2. Red1- and Mtl1-Associated Factors Differentially Affect Heterochromatin Domains

(A) ChIP-chip analysis of Red1-MYC and Mtl1-MYC distribution at heterochromatin island loci. (B) ChIP-chip analysis of H3K9me2 distribution at heterochromatin island loci in indicated strains. (C) ChIP-chip analysis of the effect of pir1Δ and rmn1Δ on H3K9me2 distribution at HOODs in rrp6Δ (top). Normalized number of small RNA reads plotted in alignment with HOOD loci (bottom). The signals above and below the line represent small RNAs that map to the top and bottom DNA strands, respectively. See also Figure S2 and Table S1.

Figure 3. MTREC Regulates Gene Expression and Targets Noncoding RNAs and premRNAs Degraded by the Exosome

(A) ChIP-chip analysis of Red1-MYC and Mtl1-MYC distribution at indicated loci (top). RNA-Seq analysis of Red1 and Mtl1 bound genes in red1Δ, mtl1-1, and rrp6Δ strains (bottom). (B) RT-PCR analysis of Red1 and Mtl1 bound loci in indicated strains. The transcript level of act1 was used as a control. +RT and −RT indicate presence or absence of reverse transcriptase. (C) ChIP-chip analysis of Red1-MYC and Mtl1-MYC distribution at snoRNAs (top). Northern analysis of snR99 and snR3 processing in mutant strains (bottom). −/+ dT denotes RNase H treatment in the absence or presence of oligo dT. A− /A+ indicates non-polyadenylated and polyadenylated species. Lighter exposures of mature snR99 and snR3 are shown below. rRNA was used as a loading control. (D) Red1-MYC and Mtl1-MYC distribution at rpl30-2 (top). Northern blot analysis of rpl30-2 pre-mRNA processing in the indicated strains (bottom). See also Table S2.

Figure 4. ncRNA Regulates Gene Expression in Response to Growth Conditions

(A) ChIPchip analysis of Red1-MYC and Mtl1-MYC distribution at the pho1 locus. (B) Schematic depicting transcription of ncRNA and mRNA at the pho1 locus (left). The location of the Northern blot probe is indicated (thick wavy line). The dotted line indicates the site cut by RNase H. Northern blot of ncRNA in wild-type and rrp6Δ strains (right). The asterisk denotes ncRNA cleaved by oligo-directed RNase H treatment. Lanes shown are from the same Northern blot with additional control lanes removed. (C) ChIP analysis of Red1-MYC and Mtl1-FLAG enrichment at the pho1 locus. Numbers shown below ChIP lanes represent fold enrichments. (D) Northern blot analysis of pho1 mRNA expression in ncRNAΔ. (E) ChIP analysis of H3K9me2 enrichment at pho1 in wild-type and ncRNAΔ. Numbers shown below ChIP lanes represent fold enrichments. (F) Northern blot analysis of ncRNA and pho1 mRNA expression in the indicated strains. See also Figure S3.

Figure 5. Mtl1 Forms Red1-independent Interactions with Nrl1 and Ctr1

(A) Expression of Nrl1-MYC and Ctr1-MYC. The arrow indicates MYC-tagged Nrl1 or Ctr1. (B) Proteins associated with Mtl1, Nrl1 or Ctr1. Proteins highly associated with each other are shaded in blue. (C) Immunofluorescence analysis of MYC-tagged Nrl1 and Ctr1. (D) Co-IP analysis of Mtl1 with Nrl1 and Ctr1. (E) Normalized number of small RNA reads plotted in alignment with HOOD 10 and HOOD 31 loci in rrp6Δ and rrp6Δ nrl1Δ strains (top). ChIP-chip analysis of H3K9me2 at HOODs (bottom). See also Figure S4 and Table S1.

Figure 6. Nrl1 Interacts with Splicing Factors That Promote HOOD Assembly at Loci Encoding Cryptic Introns

(A) Splicing factors identified in Nrl1 purifications. (B) Co-IP analysis of the interaction between Cwf10 and Nrl1. Asterisk indicates the correct size of Nrl1. Multiple Nrl1 bands may indicate protein modifications. (C) Schematic of cryptic introns detected at HOODs by RNA-Seq in indicated strains. Tf2 and SPCC1442.04c are shown with cryptic introns (brown). The arcs below the line represent intron junction reads that map to the bottom DNA strands. The thickness of the arc corresponds to the number of reads detected. The total unnormalized read count for each locus, which is indicative of sequencing depth, is shown. The grey line denotes antisense RNA. (D) Normalized number of small RNA reads at HOOD 31 in rrp6Δ and rrp6Δ 04c-int2Δ (top). ChIP analysis of H3K9me2 enrichment (bottom). Numbers shown below ChIP lanes represent fold enrichments. (E) Normalized number of small RNA reads at HOOD 31 and HOOD 10 in rrp6Δ and rrp6Δ cwf10-1 (top). ChIP-chip analysis of H3K9me2 (bottom). See also Figures S5–7 and Tables S1 and S3.

Figure 7. Ctr1 Mediates Telomere Maintenance by Regulating Telomerase RNA Processing

(A) Schematic of TER1 RNA showing the RT-PCR amplification region (top). RT-PCR analysis of splicing in indicated strains (bottom). (B) Proposed roles for different factors in TER1 RNA processing. The Sm and Ctr1-Mtl1 proteins process TER1 to the mature form. Red1-Mtl1 targets spliced TER1 for degradation by the exosome but might also inhibit complete splicing. (C) Schematic of TER1 RNA showing the probe used for Northern analysis (top). Dotted lines indicate regions cut by oligo-directed RNase H. Northern blot analysis of TER1 (bottom). A lighter exposure of mature TER1 is shown below. Relative quantitation of mature TER1 to wild-type is indicated. (D) Southern blot analysis of telomere length. (E) Schematic illustrating the role of MTREC and Mtl1-Nrl1 protein assemblies in degradation of various RNA species and formation of heterochromatin at islands or HOODs. Mtl1 in association with Ctr1 also promotes processing of telomerase RNA. Intron-containing RNA species can be targeted by the Mtl1-Nrl1 complex in association with splicing factors, or by MTREC bound to Pab2-Rmn1 and Pla1 (not shown), to degrade RNAs and assemble HOODs through recruitment of RNAi machinery, or promote RNA decay by the exosome. See also Figure S4A.